// prompts & results
Generative media with control
Prompt examples, model choices and practical outputs for images, music, video and brand assets.
The goal is not random AI art. The goal is usable media for a product, campaign or interface.
// gallery
Prompts & results
Real prompts from projects — what comes out of different AI systems.
ComfyUI · SDXLImage
1024 × 1024// prompt
"anime girl, cyberpunk city, neon orange glow, dark rainy night, ultra detailed, 8k, photorealistic"
Result: A cyberpunk-style portrait. ~45s on an RTX 4060. CFG 7, 30 steps, DPM++ SDE.
Suno AIMusic
2:34 · MP3// prompt
"happy birthday song for Olena, upbeat pop, Ukrainian, warm female vocals, acoustic guitar"
Result: A personalised song with the name. The client picked the most emotional of 3 variants.
Runway MLVideo
1080 × 1920 · 10s// prompt
"coffee cup on wooden table, slow cinematic zoom, warm morning sunlight, steam rising, 4K"
Result: A video for a café's Reels. Image-to-Video from a reference photo.
ComfyUI · FluxImage
2048 × 2048// prompt
"product mockup, white ceramic mug with custom logo, studio lighting, clean background"
Result: A mockup for a shop. Inpainting — logo without regenerating the background.
MidjourneyImage
1792 × 1024// prompt
"tech startup poster, dark background, geometric neon shapes, orange accent, minimalist --ar 16:9 --v 6"
Result: A conference poster. 4 variants in 30s, refined in Photoshop.
Kling AIAnimation
512 × 512 · GIF// prompt
"cute robot mascot waving hand, idle blinking loop, anime chibi, transparent bg, 24fps"
Result: An animated mascot for a bot. A 512px GIF sticker, ~40 frames, optimised for TGS.
// creative pipeline
Need a consistent visual system, not one lucky output?
I can build a repeatable prompt/model workflow for banners, product shots, UI moodboards, stickers or short videos.